Is this the final update from Ought about their factored cognition experiments? (I can’t seem to find anything more recent.) The reason I ask is that the experiments they reported on here do not seem very conclusive, and they talked about doing further experiments but then did not seem to give any more updates. Does anyone know the story of what happened, and what that implies about the viability of factored-cognition style alignment schemes?
Is this the final update from Ought about their factored cognition experiments? (I can’t seem to find anything more recent.) The reason I ask is that the experiments they reported on here do not seem very conclusive, and they talked about doing further experiments but then did not seem to give any more updates. Does anyone know the story of what happened, and what that implies about the viability of factored-cognition style alignment schemes?